Acoma_Andy

Put up that MCH to at least 1.26v, and keep LLC enabled at own risk.

__________________

Mattb2e

1.2mch does seem a bit low, but then again the p45 chipset runs a 65nm process if I remember correctly so the voltage need for higher fsb is less than that of the older 90nm tech of the p35.

I would imagine that you would be fine around 1.3v for your current settings, so I would try 1.26 like Acoma Andy suggested and work your way up from that. Try running your stress test and see what happens.

__________________
"Being wrong is erroneously associated with failure, when in fact to be proven wrong should be celebrated. For it is elevating someone to a new level of understanding, furthering awareness."

"Always question the recieved reality, the consensus reality is often intentionally misleading"

Thermal Compound Application Instructions:
http://www.arcticsilver.com/arctic_s...structions.htm

Intel E8400 Club

Shado

Something wierd happened. I took my lan rig out to lan party tonight, and on my way there, I made a turn too sharp and my rig fell onto it's side onto some clothes in my trunk.

I didn't see any damage on my case, but when I plugged it in, it will no longer boot at 4ghz (445x9) at the BIOS settings that was stable for 6 hours of OCCT. I tried to boot it up a couple times and then my BIOS was wiped.

I have reverted it back to 3.8 at the moment, but is it possible to have damaged something when it fell onto its side.

Shado

I updated the BIOS to F9 and everything is up and running again. I upped the MCH core to 1.26V and is now 10 hour OCCT stable. Tried to lower VCore by 2 notches down (to 1.2875) and it didn't past 1 hour OCCT.

E8500 @ 4ghz (445 x 9)
BIOS is currently set at:
LLC enabled
1.3V VCore
1.26V VTT
1.26V MCH Core
1.1V ICH Core
rest are manually set to the normal value

I tried to up the multiplier to 9.5 with the same settings and it failed to load into Windows. I guess MattB has a better chip or something than mine. What should I do if I want to try to get to 4.2Ghz? Or is it feasible to even consider that with the current settings I have? I am shooting for a 24/7 stable gaming platform.

Mattb2e

your gonna need to up your vcore and mch a bit more to get 4.2, your good until 1.36 if temps allow it.

__________________
"Being wrong is erroneously associated with failure, when in fact to be proven wrong should be celebrated. For it is elevating someone to a new level of understanding, furthering awareness."

"Always question the recieved reality, the consensus reality is often intentionally misleading"

Thermal Compound Application Instructions:
http://www.arcticsilver.com/arctic_s...structions.htm

Intel E8400 Club

Shado

argggg I'm getting frustrated here. After running the above settings in BIOS for 10 hours, I thought I had it; the stable 24/7 gaming/folding machine at 4 ghz.

I unplugged the machine and carried it downstairs, after plugging it and starting it back up, the BIOS won't boot at the setting I had that was passed 10 hours of OCCT.

What's worse, the BIOS started acting weird, won't allow me to go into BIOS for adjustment. I had to clear BIOS to get it back up. and when I entered the same settings that was stable for 10 hours, it booted.

I was skeptical and allowed the machine to fold (don't want to waste power just to stress test it again). 3 hours later when I get home, it crashed and was in the windows repair screen.

How am I supposed to tell it's stable when 10 hours of OCCT doesn't prove it?

EDIT: Looks like LLC wasn't enabled on in the BIOS the last time it crashed. Going to enable it and then let it fold. Hope I can report back with some good news.

P.S. sorry if I sound a little whiney, but my rig is starting to annoy me

2nd EDIT: I just noticed that CPUZ and CPUID Hardware monitor are both reporting VCore of 1.26 instead of 1.28 that was reported when I ran OCCT that ended up stable for 10 hours, even when BIOS was set to the same 1.30VCore. Don't know if that has anything to do with it

Mattb2e

That would be a function of Vdroop, essentially what Load Line Calibration almost completely eliminates. Thats quite a hefty Vdroop I might add. If your setting it to 1.3 and getting 1.26 in windows thats a pretty steep drop.

I dont recommend the use of LLC but thats my personal opinion, Vdroop is a function of intel's design to protect the processor from overvoltage between load and idle fluctuations, the higher the vcore whilst eliminating Vdroop, the greater the chance of damaging your cpu.

__________________
"Being wrong is erroneously associated with failure, when in fact to be proven wrong should be celebrated. For it is elevating someone to a new level of understanding, furthering awareness."

"Always question the recieved reality, the consensus reality is often intentionally misleading"

Thermal Compound Application Instructions:
http://www.arcticsilver.com/arctic_s...structions.htm

Intel E8400 Club

Mattb2e

try setting your multi to 9.5, your fsb to 422, this will give you a little over 4ghz with a smaller fsb equalling less stress on your NB. Keep the rest of the settings the same and see if that fixes the problem.

__________________
"Being wrong is erroneously associated with failure, when in fact to be proven wrong should be celebrated. For it is elevating someone to a new level of understanding, furthering awareness."

"Always question the recieved reality, the consensus reality is often intentionally misleading"

Thermal Compound Application Instructions:
http://www.arcticsilver.com/arctic_s...structions.htm

Intel E8400 Club

Shado

I did read about the LCC and how it is dangerous at high voltage, but I am running at 1.3 (still 0.06 to Intel specified max of 1.36?) so I thought it would be ok to enable it.

With LCC enabled, my system is currently stable at 1.30VCore BIOS setting (CPUZ shows 1.26-1.28), without LCC, I am unstable at 1.315VCore BIOS setting (CPUZ showing 1.25-1.28)

I'm really curios how you can get the E8400 to run at 4ghz at 1.26V, when my E8500 can barely stay stable at 1.3V with LLC

Shado

Let me give that a shot